All-wireless and, possibly, all-mobile communication networks are likely to operate in the future for both commercial and military applications, as a result of increasing pressure for communication while ``on the move.'' Such networks differ from traditional ones in fundamental ways since they may not share the fixed, ``wired'' communication infrastructure. They are refereed to as ``multihop wireless networks". They may arise as alternative means of providing personal communication services, or as specialized networks that are ``stand-alone'' or are weakly coupled to the general infrastructure. The researchers propose to study the fundamental principles of operation and design that are unique to these networks, and to develop a prototype design for a bi-level (two tier) multihop architecture that will be based on these principles.

Specifically, the researchers plan to address the following questions and to study the associated performance trade-offs that pertain mostly to layers 2-4 in the networking architecture: a) what routing algorithms are well-suited to variable connectivity environments? b) what address and mobility management methods are appropriate? how are they coupled to routing in this context? c) what multicasting methods are possible? d) what should be the role of power management as a network control parameter to conserve battery energy? e) what are the fundamental trade offs between ``push" and ``pull" communications in asymmetric network architectures? f) what is the best utilization of path diversity for reliability and efficiency under channel impairment conditions? g) what resource allocation approaches are appropriate?

By leveraging other support, the researchers intend to implement some control policies developed as a result of this investigation in an existing network testbed, and use this experience as an aid in developing a design model for multihop wireless networks. This testbed exists in the Hybrid Networks Laboratory of the Institute for Systems Research at the University of Maryland and leverages the existing equipment and human resource facilities available at the lab.

Agency
National Science Foundation (NSF)
Institute
Division of Computer and Network Systems (CNS)
Type
Standard Grant (Standard)
Application #
9818913
Program Officer
Admela Jukan
Project Start
Project End
Budget Start
1999-08-15
Budget End
2003-07-31
Support Year
Fiscal Year
1998
Total Cost
$671,044
Indirect Cost
Name
University of Maryland College Park
Department
Type
DUNS #
City
College Park
State
MD
Country
United States
Zip Code
20742